On 2/14/2024 11:50 PM, Ard Biesheuvel wrote:
I looked at the EFI spec again, and EfiACPIReclaimMemory is not
actually listed as a memory type that has this 64k alignment
requirement. This makes sense, given that this memory type has no
significance to the firmware itself, only to the OS. OTOH, reserved
memory does appear there.
So I suggest we fix that first, and then drop any mention of
EfiACPIReclaimMemory from this patch. At least we'll have heap guard
coverage for ACPI table allocations on arm64 going forward.
The logic in question was added in 2007 in commit 28a00297189c, so
this was probably the rule on Itanium, but that support is long gone.
I wanted to understand this a little bit further. I do see that the spec
does not call out EfiACPIReclaimMemory as needing 64k alignment.
However, your statement that the memory type does not have have
significance to the FW, only to the OS, so we don't need the 64k
alignment is confusing to me. Isn't the 64k alignment needed only for
regions that the OS does care about? In order to read this information
at their 64k page granularity, doesn't this need to be aligned to 64k?
Or are you saying the OS can just read the 64k page(s) containing these
4k aligned EfiACPIReclaimMemory pages and then start reading at the
calculated offset from within the larger page, since these pages don't
have any pointers to outside memory, unlike image sections?
Thanks,
Oliver
-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#115533): https://edk2.groups.io/g/devel/message/115533
Mute This Topic: https://groups.io/mt/104364784/21656
Group Owner: devel+ow...@edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
